Recent clinical results support the possibility that induction of differentiation is an alternative approach for the treatment of some malignancies. However, a greater understanding is needed of both the process of terminal differentiation and the metabolism of known inducers. To aid in this search, studies were conducted to: a) measure changes in the levels of oncogene proteins during differentiation and to correlate changes in the levels of oncogene proteins with other changes associated with differentiation; and b) study the metabolism of known inducers of differentiation. The human myeloblastoid cell line, HL60, has been a useful model system for studying terminal differentiation. A method was developed to fix and permeabilize HL60 cells so that intracellular antigens could be detected with specific antibodies and measured with a fluorescence activated cell sorter (FACS). We examined the level of c-myc oncogene protein during retinoic acid- and dimethylsulfoxide-induced differentiation and found that both the rate and the extent of decreases in c-myc protein during differentiation are much less than what would be expected from the decreases in c-myc mRNA levels under similar conditions. Although many biological effects of retinoic acid have been described the mechanism for these actions is unknown. We have now discovered that in HL60, a covalent bond is formed between retinoic acid and protein. Based on sensitivity to hydrolysis with hydroxylamine, about 80% of the retinoic acid moiety is linked to protein via either an oxygen-ester or a thio-ester bond. Fractionation of cells labeled with radioactive retinoic acid indicates that greater than 80% of the retinoylated protein is associated with the membrane fraction.